The business challenge
Pricing approximately 1,200 generic pharmaceutical SKUs required balancing demand, customer differences, and competitive pressure. Quotes took five days, and large-account bids had a 31% win rate. Earlier visibility into potential competitor entry was another opportunity to improve pricing decisions.
The architecture behind the outcome
I architected a segment-aware pricing platform serving large, medium, and small customers. Demand and elasticity models work alongside bid win-probability models and Gurobi optimization. LLM-written explanations surface the rationale inside Salesforce CPQ, making recommendations usable within the existing sales workflow.
Delivery & decision ownership
- Model demand, elasticity, and bid win probability with customer segmentation.
- Combine predictive outputs with Gurobi optimization to support pricing recommendations.
- Use FDA generic drug approval data as an early warning of price erosion, one to two quarters before competitor entry.
- Present LLM-written explanations directly inside Salesforce CPQ.
Results that matter
| Measure | Before | After / outcome |
|---|---|---|
| Gross margin | Baseline | +2.6 percentage points |
| Large-account bid win rate | 31% | 42% |
| Quote turnaround | 5 days | 6 hours |
| Competitive early warning | — | 1–2 quarters |
| Payback period | — | 6 months |
The pricing program delivered $15.0M in annual benefit with a six-month payback. Commercial teams gained faster quote turnaround, stronger bid performance, and earlier signals of competitive price erosion.